Variable Percentage Withdrawal (VPW)

[ch4au2]

If I'm the only one that is confused, maybe someone else on this thread can take a crack at an explanation that doesn't re-repeat the one I quoted originally.

Unfortunately I'm not going to be of any help. I cannot make you understand what you don't understand.

I'll just say that, IMHO, all these withdrawal strategies and calculations are so speculative that handwringing over the details is not worth it. Just take it as another data point put together by people who have a strong grasp on all these things.

What I did for the first 13 years of retirement, was use my own spreadsheet which has all the normal SS, pension, etc. cells to put in and a cell for "return" which is just a nominal percentage I can choose. For the past 13 years, I've put in 2%, then I've also not withdrawn what that says I can withdraw. Both of those being the conservative bogleheads type of thing to do. The combination of making a lot higher returns (average of something like 8%), and not taking out what the spreadsheet has calculated, has gotten me in a wonderful position! One that becomes more and more comfortable, but one where I want to start taking more and enjoying a few more perks.

What I am now doing is using every calculation like this I can find, and see how close they are in what they say you can withdraw. Ficalc.com is a great online resource and they have something like 12 different withdrawal strategies you can choose from using a drop-down menu, including VPW. In addition, you can tweak each strategy using things like "minimum withdrawal" in any given year and you can also set a "maximum withdrawal."

What I'm doing is looking at all these methods, or at least the ones that satisfy my requirements (I'm not interested in an endowment strategy), and trying to understand what makes their amounts different. And then logically asking myself the question of "what makes the low ones low and what makes the high ones high" and does either of those make any sense to me in my situation. And at then end of the day coming up with a plan that is on the relatively conservative end of things, but not as conservative as I've been doing.

I can attest that taking out less than you can is both good and bad. I've done that for 13 years. It's good because it keeps you safe and comfortable. It's bad because at some point you realize that many of the things you might have liked to do, but didn't because your stiff and conservative.. I should have done! But we can only go forward, so I'm trying to "loosen up a bit" and learn how to spend a bit more!

[Wrench]

If I'm the only one that is confused, maybe someone else on this thread can take a crack at an explanation that doesn't re-repeat the one I quoted originally.

OK - I'll take a shot. The VPW withdrawal percentages are just fixed numbers in a table. They are a given. Period. If you don't like them, then use another withdrawal approach. If you want to know how they are calculated, i.e., "look behind the curtain" so to speak, then they are based on stock and bond returns that are reasonable based upon history and some other assumptions built in like having an income floor at age 100 so withdrawals can end there, and using your entire portfolio balance during your lifetime. There's a whole thread on the details:
viewtopic.php?p=7471071#p7471071
but again if you use VPW, you take the results as a given. The very good news is, you never run out of money up to age 100 and withdrawal percentages can be higher than some other approaches. You "pay" for that benefit by accepting variable withdrawals every year. Bottom line (and this may be unsatisfying to you): you accept it as it is. Longinvest and others have put countless hours into developing it, and at some level you just have to say OK. If you are satisfied it will work well enough for your situation, use it. If not, use some other approach. FWIW, after studying it intensely for a couple years, I decided it was not for me. I don't want the variable income, and I don't need or want to spend my entire portfolio. (And I'm planning to live to 110 so want more than an income floor at 100! ) It's an awesome tool and spreadsheet - it just doesn't fit my personal situation. YMMV.

Wrench

[tennisplyr]

I'm seeing dramatically different withdrawal rates here on this site based on age/allocation, could someone point me in the right direction. Thanks.

[LadyGeek]

A post that was moved here from the VPW forward test thread has been moved back. My mistake. Sorry about that. See: Re: A Simple Bogleheads Retirement Using Variable Percentage Withdrawals (VPW Forward Test)

[Siaigi]

As soon as the pension starts (when the youngest is 67.5 years old), though, the pension must be marked as already started. To eliminate the error resulting from the retiree's age (youngest spouse) being 67 with a pension marked as already started with a start age of 68 (after rounding up), the pension start age should be removed (empty the yellow cell).

Finally we arrived here.

Following all your recommendations we don't have anymore errors
but there is one last doubt.

The suggested annual portfolio withdrawal is now about 1/10 than before. how can it be?

[Raspberry-503]

If I'm the only one that is confused, maybe someone else on this thread can take a crack at an explanation that doesn't re-repeat the one I quoted originally.

OK - I'll take a shot. The VPW withdrawal percentages are just fixed numbers in a table. They are a given. Period. If you don't like them, then use another withdrawal approach. If you want to know how they are calculated, i.e., "look behind the curtain" so to speak, then they are based on stock and bond returns that are reasonable based upon history and some other assumptions built in like having an income floor at age 100 so withdrawals can end there, and using your entire portfolio balance during your lifetime. There's a whole thread on the details:
viewtopic.php?p=7471071#p7471071
but again if you use VPW, you take the results as a given. The very good news is, you never run out of money up to age 100 and withdrawal percentages can be higher than some other approaches. You "pay" for that benefit by accepting variable withdrawals every year. Bottom line (and this may be unsatisfying to you): you accept it as it is. Longinvest and others have put countless hours into developing it, and at some level you just have to say OK. If you are satisfied it will work well enough for your situation, use it. If not, use some other approach. FWIW, after studying it intensely for a couple years, I decided it was not for me. I don't want the variable income, and I don't need or want to spend my entire portfolio. (And I'm planning to live to 110 so want more than an income floor at 100! ) It's an awesome tool and spreadsheet - it just doesn't fit my personal situation. YMMV.

Wrench

"Trust what you don't understand" goes against the grain. I understand how the math of VPW works, the need for an annuity to protect about early depletion,etc.... All variable methods have varying withdrawal based on portfolio fluctuation, and VPW uses an amortization table to decide how much to withdraw in each year. In that regard it is identical to ABW, TPAW, possibly RMDs. I'm still working through it but it seems that both TPAW and VPW use very similar ways to account for the PV of SS and future pension, etc... TPAW adds another layer by using PV of future spending, and guides you through choosing the AA that best matches those goals (at the expense of upside), but at its simplest both methods work basically the same way, and give very similar results for a given growth trend/expected results. So it seemed that the main difference remains in whether those are immutable or not, and I was trying to understand the reasoning. Responses of "don't worry about it", "take it or leave it" or "it's too hard to guess future rates" don't quite scratch that itch.
Again I'm not criticizing, but trying to understand choices by the many people over long hours. I'm OK with saying VPW is "good enough" in its simplicity, but there seems to be such an allergic response to the idea of even mentioning expected returns, that I was trying to understand it.

I'll stop asking.

[longinvest]

As soon as the pension starts (when the youngest is 67.5 years old), though, the pension must be marked as already started. To eliminate the error resulting from the retiree's age (youngest spouse) being 67 with a pension marked as already started with a start age of 68 (after rounding up), the pension start age should be removed (empty the yellow cell).

Finally we arrived here.

Following all your recommendations we don't have anymore errors
but there is one last doubt.

The suggested annual portfolio withdrawal is now about 1/10 than before. how can it be?

Siaigi, normally, the amount available for taxes and expenses (which includes the portfolio withdrawal and pension payments) should be relatively similar to what it was before. Is that the case?

Generally, it's normal for the withdrawal to get smaller when a pension starts because the worksheet stops including a bridging amount for the pension as part of the withdrawal. Also, if the pension doesn't have cost-of-living adjustments, the withdrawal is further reduced to dampen the erosive impact of inflation on the purchase power of pension payments.

[Siaigi]

Is that the case?

Yes it is my case! Almost the same amount.

Please, where could I read more about the bridging amount?

[silent cal]

[Thread merged into here --admin LadyGeek]

Hi everyone,

I have some questions about using the VPW Accumulation Worksheet as a married couple and am curious to hear your thoughts?

Husband
Age: 25
Annual Salary: ($30,000 + $1,500 employer match) = $31,500
Portfolio Balance: $60,000
Portfolio Allocation: 90/10 stocks/bonds
Retirement Age: 50
Portfolio Contribution Frequency: Biweekly

Defined Benefit Pension #1

Name: Husband Social Security
Start Age: 70
Monthly Payment: $1,670 (75% haircut of my personalized SSA estimate, per my opinion on the future of SS)
Cost of Living Adjustments: Yes
Annual Contribution: $2,295 (calculation excludes employer 401(k) match included in salary above)

Wife
Age: 25
Annual Salary: ($61,000 + $5,063 employer match) = $66,063
Portfolio Balance: $0
Portfolio Allocation: 90/10 stocks/bonds
Retirement Age: 50
Portfolio Contribution Frequency: Biweekly

Defined Benefit Pension #2

Name: Wife Social Security
Start Age: 70
Monthly Payment: $2,146 (75% haircut of my personalized SSA estimate, per my opinion on the future of SS)
Cost of Living Adjustments: Yes
Annual Contribution: $4,666 (calculation excludes employer 401(k) match included in salary above)

Here are my questions:

• Do we list each of our Social Security contributions as separate Defined Benefit pensions? If I recall on the Retirement spreadsheet it is recommended not include the lower SS amount. Should we likewise do that here?

• We are currently invested aggressively in a 90/10 portfolio and plan to rebalance towards a 60/40 portfolio at retirement. As long as we update this spreadsheet annually to show changes to our AA, the portfolio balance target amount should likewise adjust right? That is to say, we aren't 'messing up' the calculations if we are currently a 90/10 but plan to be at a 60/40 AA when we use the Retirement sheet?

• Should the employer match be included as part of annual salary? Then, to calculate our 401k target amounts, we would subtract our biweekly employer matches from the 'Biweekly Portfolio Contribution' cell to determine what we need to save outside of the match?

• Is the idea that the 'Required Flexibility' section in the Accumulator spreadsheet is essentially designed to mimic the 'Required Flexibility' in the Retirement spreadsheet? So that as an Accumulator approaches and then enters Retirement, if they were able to find flexibility in their budget as an Accumulator, they can roughly expect the Retirement flexibility dollar amount to be similar? Say in the last year as an Accumulator vs first year as a Retiree?

[silent cal]

Does anyone know how VPW works with Roth conversions? I'm struggling conceptually to understand how they impact filling in the portfolio balance and what your recommended withdrawal should be?

[azanon]

Does anyone know how VPW works with Roth conversions? I'm struggling conceptually to understand how they impact filling in the portfolio balance and what your recommended withdrawal should be?

I only know that VPW is based on pre-tax money. To value Roth money, I guess you'd have to add back in whatever your combined marginal rates are. And since you asked about Roth conversions, the figure that goes in VPW is before you convert it, so again pre-tax.

Remember, VPW goal is to figure out a smooth amount available for TAXES and expenses.

[LadyGeek]

I merged silent cal's thread into the ongoing support thread.

(Thanks to the member who reported the post and provided a link to this thread.)

[RyeBourbon]

Does anyone know how VPW works with Roth conversions? I'm struggling conceptually to understand how they impact filling in the portfolio balance and what your recommended withdrawal should be?

I only know that VPW is based on pre-tax money.

Why do you say that?

[davidski]

Hello, all. I'm looking to apply the VPW spreadsheet in a post-financial independence, pre-retirement, scenario. Referencing this advice from longinvest:

Once the target financial independence age is reached, it's time to switch to the Retirement sheet and start practicing retirement by filling it with age, portfolio balance, and accrued pension promises (excluding pension increases due to continuing to work), while ignoring the continuing salary. Portfolio contributions (or withdrawals) become the difference between salary and suggested portfolio withdrawal. This will result into a generally increasing amount available for taxes and expenses. At any point in time, the investor can decide to stop working and retire for real, thanks to financial independence.

When looking at the resulting VPW annual income (and required flexibility), taking those amounts minus gross salary gives a contribution amount many times the amount on the accumulation tab. Should the salary being bench-marked against be the net salary (gross - non retirement withholding)?

As always, a big thanks to longinvest and the rest of the VPW community for the tool and the surrounding conversations!

[azanon]

Does anyone know how VPW works with Roth conversions? I'm struggling conceptually to understand how they impact filling in the portfolio balance and what your recommended withdrawal should be?

I only know that VPW is based on pre-tax money.

Why do you say that?

VPW is attempting to mathematically match the money available for taxes and expenses before retirement to that after retirement. The only way it can do that accurately, is if the inputs being used haven't been taxed yet. The standard inputs are social security, pension, and either earned income or tax advantaged retirement accounts - read all monies that are subject to income tax.

[RyeBourbon]

Does anyone know how VPW works with Roth conversions? I'm struggling conceptually to understand how they impact filling in the portfolio balance and what your recommended withdrawal should be?

I only know that VPW is based on pre-tax money.

Why do you say that?

VPW is attempting to mathematically match the money available for taxes and expenses before retirement to that after retirement. The only way it can do that accurately, is if the inputs being used haven't been taxed yet. The standard inputs are social security, pension, and either earned income or tax advantaged retirement accounts - read all monies that are subject to income tax.

That doesn't make sense to me. It uses the entire portfolio and tells you what to withdraw. Some of it you will use to pay taxes. VPW doesn't know or care whether you withdraw from pre-tax or otherwise. That's up to you to figure out. I use taxable, tax deferred and Roth to target a particular AGI and the tax is just another expense.

[azanon]

Does anyone know how VPW works with Roth conversions? I'm struggling conceptually to understand how they impact filling in the portfolio balance and what your recommended withdrawal should be?

I only know that VPW is based on pre-tax money.

Why do you say that?

VPW is attempting to mathematically match the money available for taxes and expenses before retirement to that after retirement. The only way it can do that accurately, is if the inputs being used haven't been taxed yet. The standard inputs are social security, pension, and either earned income or tax advantaged retirement accounts - read all monies that are subject to income tax.

That doesn't make sense to me. It uses the entire portfolio and tells you what to withdraw. Some of it you will use to pay taxes. VPW doesn't know or care whether you withdraw from pre-tax or otherwise. That's up to you to figure out. I use taxable, tax deferred and Roth to target a particular AGI and the tax is just another expense.

My apologies for not explaining it well enough. I have no doubt about what i said. Perhaps longinvest can confirm.

[Raspberry-503]

So if you were to do large conversions on a smallish portfolio, the portfolio balance will go down some due to the conversion taxes and impact the next years withdrawal
Presumably the impact would remain small, and your effective tax rate in the future be less and let you keep more of future withdrawals

[longinvest]

Yes it is my case! Almost the same amount.

Please, where could I read more about the bridging amount?

Siaigi, I'll explain this with an example.

Withdrawals are annual and, for simplicity, all amounts are in constant 2024 dollars.

The retiree is 65 years old, has a $1,000,000 portfolio with a 60/40 stocks/bonds allocation, and is delaying Social Security to start receiving $3,000/month at age 70. That's a (12 X $3,000) = $36,000/year missing income for the first 5 years of retirement.

To compensate for this missing income at age 65 the VPW worksheet divides the missing annual income by the 5-year withdrawal percentage (21.5%) for a 60/40 stocks/bonds portfolio in the VPW Table, resulting into ($36,000 / 21.5%) = $167,442. Then, the VPW worksheet divides, on paper, the $1,000,000 portfolio into two separate portfolios, a $167,442 portfolio from which 21.5% is extracted to replace missing Social Security payments for the current year, and a remaining $832,558 portfolio from which a normal 5.0% (age 65, 60/40 stocks/bond allocation) VPW withdrawal is taken.

As a result, at age 65, the retiree takes a (($167,442 X 21.5%) + ($832,558 X 5.0%)) = ($36,000 + $41,628) = $77,628 portfolio withdrawal.

There's a single $1,000,000 portfolio, not two portfolios. The division is only made temporarily on paper to calculate the portfolio withdrawal. After withdrawal the remaining balance is $922,372. During the year, the portfolio is exposed to market fluctuations. Let's assume that, one year later, the portfolio has increased by 5% to ($922,372 X (1 + 5%)) = $968,491 and the retiree is now 66 years old.

At age 66, the worksheet makes a similar calculation, dividing $36,000 by the 4-year withdrawal percentage (26.4%) for a 60/40 stocks/bonds portfolio. The retiree ends up taking a (($136,364 X 26.4%) + ($832,127 X 5.1%)) = ($36,000 + $42,438) = $78,438 portfolio withdrawal.

This process is repeated at ages 67, 68, and 69.

Finally, at age 70, the retiree starts receiving $3,000/month Social Security payments. Let's assume that the portfolio is $775,000. The worksheet applies the 5.4% withdrawal percentage for age 70 and a 60/40 stocks/bonds portfolio in the VPW Table to the entire portfolio, resulting into a ($775,000 X 5.4%) = $41,850 portfolio withdrawal. Combined with $36,000 from Social Security, the retiree gets $77,850 in total annual retirement income, similar to earlier years, despite taking a significantly smaller portfolio withdrawal.

I suggest taking the time to read the forward test thread, including monthly posts like this one I made today, explaining similar calculations displayed in the bottom part of the Retirement sheet.

[longinvest]

Hi everyone,

I have some questions about using the VPW Accumulation Worksheet as a married couple and am curious to hear your thoughts?

Husband
Age: 25
Annual Salary: ($30,000 + $1,500 employer match) = $31,500
Portfolio Balance: $60,000
Portfolio Allocation: 90/10 stocks/bonds
Retirement Age: 50
Portfolio Contribution Frequency: Biweekly

Defined Benefit Pension #1

Name: Husband Social Security
Start Age: 70
Monthly Payment: $1,670 (75% haircut of my personalized SSA estimate, per my opinion on the future of SS)
Cost of Living Adjustments: Yes
Annual Contribution: $2,295 (calculation excludes employer 401(k) match included in salary above)

Wife
Age: 25
Annual Salary: ($61,000 + $5,063 employer match) = $66,063
Portfolio Balance: $0
Portfolio Allocation: 90/10 stocks/bonds
Retirement Age: 50
Portfolio Contribution Frequency: Biweekly

Defined Benefit Pension #2

Name: Wife Social Security
Start Age: 70
Monthly Payment: $2,146 (75% haircut of my personalized SSA estimate, per my opinion on the future of SS)
Cost of Living Adjustments: Yes
Annual Contribution: $4,666 (calculation excludes employer 401(k) match included in salary above)

Here are my questions:

• Do we list each of our Social Security contributions as separate Defined Benefit pensions? If I recall on the Retirement spreadsheet it is recommended not include the lower SS amount. Should we likewise do that here?

Silent cal, there's a variety of approaches to handle finances in a couple: fully combined, separate, hybrid. This can have an impact on retirement planning.

There's no specific recommendation for couples in the Retirement sheet of the VPW worksheet, as far as I know.

But, earlier in this thread, I've discussed this topic. What I've explained is that it's important to consider what will happen after first death. As a result, some couple might decide to only include part of their pensions in their planning to dampen the financial impact of first death for the survivor. Not all couples will want to do this. For various reasons, some couples prefer to maintain separate finances. For example, this can happen when both spouses have their own children from previous relations.

• We are currently invested aggressively in a 90/10 portfolio and plan to rebalance towards a 60/40 portfolio at retirement. As long as we update this spreadsheet annually to show changes to our AA, the portfolio balance target amount should likewise adjust right? That is to say, we aren't 'messing up' the calculations if we are currently a 90/10 but plan to be at a 60/40 AA when we use the Retirement sheet?

It's fine to use the worksheet with a gliding allocation. A higher stock allocation allows for a higher immediate amount available for taxes and expenses, accompanied with the possibility of a lower amount available for taxes and expenses in case of a market downturn. Mainly, what's important is to meet the required flexibility of the current allocation at all times. In early accumulation days, when the portfolio is tiny, asset allocation doesn't make much of a difference on the required flexibility. (It doesn't make much of a difference on portfolio growth, either, because portfolio contributions dominate portfolio growth).

• Should the employer match be included as part of annual salary? Then, to calculate our 401k target amounts, we would subtract our biweekly employer matches from the 'Biweekly Portfolio Contribution' cell to determine what we need to save outside of the match?

Yes.

• Is the idea that the 'Required Flexibility' section in the Accumulator spreadsheet is essentially designed to mimic the 'Required Flexibility' in the Retirement spreadsheet? So that as an Accumulator approaches and then enters Retirement, if they were able to find flexibility in their budget as an Accumulator, they can roughly expect the Retirement flexibility dollar amount to be similar? Say in the last year as an Accumulator vs first year as a Retiree?

Yes, it's the same idea, but it applies during accumulation. When the portfolio shrinks, portfolio contributions are increased, leaving less for taxes and expenses.

[woodstockga2]

Hello, I am currently using excel to find the yearly spending amount for VPW (for a few reasons, I do not want to use the official spreadsheet)

I currently have the below formula that I have been working off of.

=K4+(K5-(K4/ pmt(0.05,2060-K2,-1,0,1)))*pmt(0.05,2075-K2,-1,0,1)

Field explaination
K4 is my expected social security amount (at age 70)
K5 is my current portfolio
K2 is the current year
2075 is how long I want this money to last
The rate of 5% is my expected real growth rate (all stock)

My question is, how would you adjust this formula to include a payment of $18,000 (real) per year for the next 13 years

Thank you!

[Raspberry-503]

trying to understand your formula first

(K4/ pmt(0.05,2060-K2,-1,0,1)) is present value of your SS, or in other words how much you need to have in year K2 to pay K4 dollars yearly until SS taxes over in 2060 ? (could have been =-PV(0.05,2062-K2,K4,0)? they're close enough)

so your portfolio minus that amount is what you get to withdraw from at the VPW rate.

since the VPW table doesn't allow for 100% stock allocation, you calculate it yourself with pmt(0.05,2075-K2,-1,0,1)

I think it makes sense to me. so now you need to also discount your current portfolio by the present value of those future payments, PV(5%, 13, 18000)

=K4+18000+(K5-(K4/ pmt(0.05,2060-K2,-1,0,1))-pv(5%,13,18000) )*pmt(0.05,2075-K2,-1,0,1)

you'll need to adjust the 13 in the formula for years remaining after K2, I'm not clear if the starting year is 2024
you'll also need to be careful about how the formula behaves as you reach then od the 13 years and start of SS at 2060 as some of the terms of the formula need to disappear. I think that's right off the top of my head. that would be one way to say you want to spend an extra $18K for travel for the first 13 years of retirement. That factor wouldn't scale for market conditions though, it still assumes you withdraw $18K in future years after a crash.

Edited, sorry bad cut and paste in my original post so the formula was wrong.
Edited again to add the 18K on the front. the formula changes once you are past the first 13 years and you no longer spend the $18K

[longinvest]

Hello, all. I'm looking to apply the VPW spreadsheet in a post-financial independence, pre-retirement, scenario. Referencing this advice from longinvest:

Once the target financial independence age is reached, it's time to switch to the Retirement sheet and start practicing retirement by filling it with age, portfolio balance, and accrued pension promises (excluding pension increases due to continuing to work), while ignoring the continuing salary. Portfolio contributions (or withdrawals) become the difference between salary and suggested portfolio withdrawal. This will result into a generally increasing amount available for taxes and expenses. At any point in time, the investor can decide to stop working and retire for real, thanks to financial independence.

When looking at the resulting VPW annual income (and required flexibility), taking those amounts minus gross salary gives a contribution amount many times the amount on the accumulation tab. Should the salary being bench-marked against be the net salary (gross - non retirement withholding)?

As always, a big thanks to longinvest and the rest of the VPW community for the tool and the surrounding conversations!

Davidski, the trick, for simplicity, is to operate on an after-tax basis.

Here's an example. The accumulating investor is 55 years old, has a $1,000,000 portfolio, 50/50 stocks/bonds allocation, and would get $3,000/month Social Security payments starting at age 70 if stopping work today. We're assuming that the investor has the financial ability to retire today but chooses to continue working and start practicing retirement.

I haven't included the investor's salary (gross and net amounts). I'll do that later.

I put the above amounts into the Retirement sheet of the VPW worksheet with a monthly frequency and get the following assessment:

• Monthly income for 2024 -- Available for taxes and expenses: $5,022
• Required flexibility -- Monthly income after loss: $3,945

These are gross amounts (pre-tax). To practice retirement, the investor has to operate on an after-tax basis as explained in this post and this post. Assuming that the investor lives in California, I use the smartasset.com calculator and I enter (12 X $5,022) = $60,264 as "Annual Retirement Account Income". This gives me a ($5,364 federal + $2,147 state) = $7,511 estimate for taxes. This results into a (($60,264 - $7,511) / 12) = $4,396 monthly amount available for expenses (after estimated taxes). I leave as an exercise the required flexibility calculations which result into a -$870 reduction (after taxes).

Maybe the investor currently has a $5,000/month take home salary, but having chosen to be financially independent, the investor is willing to live on $4,396 (and possibly lower this by -$870 to $3,526 if markets misbehave). Consequently, the accumulating investor puts $4,396 of the take home salary into a spending account and invests the remaining ($5,000 - $4,396) = $604 into the portfolio.

The investor's gross salary is bigger than the $5,000 take home amount. Paycheck deductions are taken by the employer for Social Security and 401(k) contributions. As the investor is practicing retirement, these amounts aren't accounted for in calculations. They'll eventually affect the age 70 Social Security estimate (next year, as estimating future Social Security payments once a year is good enough) and the portfolio size in the worksheet (next month, as the portfolio balance will be affected by new 401(k) contributions). The additional $604 investment will also affect portfolio size next month.

My wife and I have operated similarly for a few years before retirement (this year). It was interesting to theoretically live off our portfolio while effectively living off part of our salaries and investing the excess over what our portfolio (and future pensions) allowed us to spend (without taking our ongoing salaries and future contributions into account). We were financially independent. Stepping into retirement, this year, was very smooth, as we were already managing finances (on an after-tax basis) as if retired.

[davidski]

Thank you for the detailed response! That was very helpful, particularly the reminder that Social Security and 401(k) deductions, since going to future payments, can/should be ignored for this pre-retirement check. I'm a bit excited/nervous to see that my baseline numbers are looking very solid and even my 50% loss numbers are comfortable!

[TitaniumCranium]

I apologize if this has already been covered.

It seems to me that the backtesting spreadsheet doesn't take into consideration the timing of when social security benefits will start - it assumes right away in year 1 - whereas the VPW spreadsheet does. It looks like this causes the first year's withdrawal and total income to differ between the two spreadsheets.

In the unlikely event that my observations are valid, are there any plans on modifying the backtesting spreadsheet to handle SS like the main VPW spreadsheet does?

[longinvest]

I apologize if this has already been covered.

It seems to me that the backtesting spreadsheet doesn't take into consideration the timing of when social security benefits will start - it assumes right away in year 1 - whereas the VPW spreadsheet does. It looks like this causes the first year's withdrawal and total income to differ between the two spreadsheets.

In the unlikely event that my observations are valid, are there any plans on modifying the backtesting spreadsheet to handle SS like the main VPW spreadsheet does?

TitaniumCranium, there's an answer in the last 3 paragraphs of this earlier post. I might eventually reconsider my decision, but the main problem leading me to resist doing it (widespread misuse of backtesting) remains. Anyway, as explained in the linked post, the VPW worksheet is a much better planning and stress testing tool.

[TitaniumCranium]

I apologize if this has already been covered.

It seems to me that the backtesting spreadsheet doesn't take into consideration the timing of when social security benefits will start - it assumes right away in year 1 - whereas the VPW spreadsheet does. It looks like this causes the first year's withdrawal and total income to differ between the two spreadsheets.

In the unlikely event that my observations are valid, are there any plans on modifying the backtesting spreadsheet to handle SS like the main VPW spreadsheet does?

TitaniumCranium, there's an answer in the last 3 paragraphs of this earlier post. I might eventually reconsider my decision, but the main problem leading me to resist doing it (widespread misuse of backtesting) remains. Anyway, as explained in the linked post, the VPW worksheet is a much better planning and stress testing tool.

Thank you very much, indeed.

[Raspberry-503]

If you believe in the concept of "gogo years" you could also mentally split your money into "gogo years fund" and "rest of our lives" fund. For example you may have enough to decide that you want to draw down 25% of your portfolio in the next 10 years before you are 75. you can enter the leftover 75% balance in the sheet with your real ages, etc.. for your "regular" spending, then but enter the amount dedicated to gogo in separate sheet where you say you are 90 (10 years to depletion at 100) and draw from that fund at the accelerated schedule

Here's an example from my own spreadsheet that shows how the gogo-pool (yellow) is depleted in addition to the regular VPW pool (orange). the shades of gray are SS/bridge ladder to establish a spending floor, your numbers/ratios would vary but it gives an example of how in 1960 pictured here, you were able to spend extra before age 75. Of course this being during a tough time, the pool available at age 76 is reduced, so you need to be OK with the amount available till age 83, after which the economy recovered enough to grow your spendable money again. Most retirement years look much better than that, because the portfolio/inflation are in better shape than retiring in this example of 1960. BTW in this example I did use 25% of initial balance dedicated to gogo years. I forgot to adjust the spreadsheet for starting at 65 years (your case), so it illustrates retiring at 57 (even more gogo years)

As someone pointed out, you kind of have to keep track of 2 pools of money the way I described it above. Instead I worked out my own personal withdrawal table in excel: I decided that 20% of the portfolio at age 55 should be dedicated to gogo years. So I ran calculation of how to run 20% of the portfolio out of money in the first 20 years (gogo ends at 75 in my model) and consume the remaining 80% "normally" according to the VPW table,
When I combine both I get a my own table that has more aggressive withdrawals for the first 20 years:



As long as I'm happy with that consumption pattern, I can apply those percentages to the portfolio every month to determine my monthly withdrawal

For a nominal $1M portfolio, the withdrawal before age 75 is $4,533/month and $3,301 after age 75 or 27% lower than it would have been without the increased yearly consumption.
For a $1M portfolio the monthly withdrawal without this change would be $4127, or another way to look at it is that the gogo withdrawal is 10% higher for the 1st 20 years than it would have been with straight VPW.

Those number work for me, the extra 10% early is a nice trip every year (for an average market, of course when the market crashes the withdrawals scale accordingly). The lower percentage later in life still seems comfortable based on what I know today. Remember this is mostly discretionary income on top of SS and other "guaranteed" income.

One could play with a different initial pool allocated to gogo years to suit their needs. Of course I have no idea when my body will tell me gogo is turning into slowgo, 75 is a guess based on observation of various family members,

I'm still playing with this. The amortization formulas allow for skewing consumption early or late but they use a smooth transition, not a step function before/after gogo, so it's a little harder to get more early while having enough later, but I'm still playing with it.

[Ckprocker]

If you believe in the concept of "gogo years" you could also mentally split your money into "gogo years fund" and "rest of our lives" fund. For example you may have enough to decide that you want to draw down 25% of your portfolio in the next 10 years before you are 75. you can enter the leftover 75% balance in the sheet with your real ages, etc.. for your "regular" spending, then but enter the amount dedicated to gogo in separate sheet where you say you are 90 (10 years to depletion at 100) and draw from that fund at the accelerated schedule

Here's an example from my own spreadsheet that shows how the gogo-pool (yellow) is depleted in addition to the regular VPW pool (orange). the shades of gray are SS/bridge ladder to establish a spending floor, your numbers/ratios would vary but it gives an example of how in 1960 pictured here, you were able to spend extra before age 75. Of course this being during a tough time, the pool available at age 76 is reduced, so you need to be OK with the amount available till age 83, after which the economy recovered enough to grow your spendable money again. Most retirement years look much better than that, because the portfolio/inflation are in better shape than retiring in this example of 1960. BTW in this example I did use 25% of initial balance dedicated to gogo years. I forgot to adjust the spreadsheet for starting at 65 years (your case), so it illustrates retiring at 57 (even more gogo years)

As someone pointed out, you kind of have to keep track of 2 pools of money the way I described it above. Instead I worked out my own personal withdrawal table in excel: I decided that 20% of the portfolio at age 55 should be dedicated to gogo years. So I ran calculation of how to run 20% of the portfolio out of money in the first 20 years (gogo ends at 75 in my model) and consume the remaining 80% "normally" according to the VPW table,
When I combine both I get a my own table that has more aggressive withdrawals for the first 20 years:



As long as I'm happy with that consumption pattern, I can apply those percentages to the portfolio every month to determine my monthly withdrawal

For a nominal $1M portfolio, the withdrawal before age 75 is $4,533/month and $3,301 after age 75 or 27% lower than it would have been without the increased yearly consumption.
For a $1M portfolio the monthly withdrawal without this change would be $4127, or another way to look at it is that the gogo withdrawal is 10% higher for the 1st 20 years than it would have been with straight VPW.

Those number work for me, the extra 10% early is a nice trip every year (for an average market, of course when the market crashes the withdrawals scale accordingly). The lower percentage later in life still seems comfortable based on what I know today. Remember this is mostly discretionary income on top of SS and other "guaranteed" income.

One could play with a different initial pool allocated to gogo years to suit their needs. Of course I have no idea when my body will tell me gogo is turning into slowgo, 75 is a guess based on observation of various family members,

I'm still playing with this. The amortization formulas allow for skewing consumption early or late but they use a smooth transition, not a step function before/after gogo, so it's a little harder to get more early while having enough later, but I'm still playing with it.

Thanks for the explanation. I like the concept. For those who have a lot of discretionary spending, it makes sense as long as a retiree realizes they will not have as much later in life. But for most people in retirement spending goes down.
I need to work on the outcomes for myself. No kids etc and a good amount of discretionary spending saved up to start retirement very soon in a few months.

[silent cal]

Apologies if this was answered previously, but is there a 'recommended' withdrawal strategy to use in tandem with VPW? I know the usual Boglehead suggestion is taxable > tax-deferred > tax-free. Would you simply take the recommended VPW withdrawal amount and pull it from whichever account you deem most 'tax-efficient'?

[GaryA505]

Apologies if this was answered previously, but is there a 'recommended' withdrawal strategy to use in tandem with VPW? I know the usual Boglehead suggestion is taxable > tax-deferred > tax-free. Would you simply take the recommended VPW withdrawal amount and pull it from whichever account you deem most 'tax-efficient'?

That's what I would do if I wasn't already into RMD age. As it is, I have to take the RMD first, then take the remainder from taxable. Since my wife is much younger than me and my taxable can be spend down to 0 by my expiration date, I'll probably just take the RMD from IRA and a fixed percentage (say 6% or 7%) from taxable). Then again, I'll have two kids in college soon so all bets are off!

[longinvest]

Apologies if this was answered previously, but is there a 'recommended' withdrawal strategy to use in tandem with VPW? I know the usual Boglehead suggestion is taxable > tax-deferred > tax-free. Would you simply take the recommended VPW withdrawal amount and pull it from whichever account you deem most 'tax-efficient'?

Silent cal, I suggest reading this earlier post for explanations and a simple suggestion.

[silent cal]

Apologies if this was answered previously, but is there a 'recommended' withdrawal strategy to use in tandem with VPW? I know the usual Boglehead suggestion is taxable > tax-deferred > tax-free. Would you simply take the recommended VPW withdrawal amount and pull it from whichever account you deem most 'tax-efficient'?

Silent cal, I suggest reading this earlier post for explanations and a simple suggestion.

Very helpful thank you.

I have another unrelated question. If you’re not planning on receiving any income in retirement outside of your portfolio, and in light of inflation-adjusted SPIAs going away, would you still recommend VPW?

Or perhaps a very conservative SWR instead?

[longinvest]

Very helpful thank you.

I have another unrelated question. If you’re not planning on receiving any income in retirement outside of your portfolio, and in light of inflation-adjusted SPIAs going away, would you still recommend VPW?

Or perhaps a very conservative SWR instead?

Silent cal, I don't understand the context of your question.

According to an earlier post, your couple will be getting $45,000/year (after haircut) from Social Security. Why would you ignore this in your planning?

The wiki page says:

VPW is best used in conjunction with guaranteed base income from Social Security, a pension (if any), and (if necessary) a SPIA (possibly with annual cost-of-living adjustments).

The key word is "possibly". There's no requirement for cost-of-living adjustments. The VPW worksheet takes into account pensions and SPIAs with and without cost-of-living adjustments. Maybe you had read an earlier version of the wiki page. The change was discussed in October 2023.

[silent cal]

Very helpful thank you.

I have another unrelated question. If you’re not planning on receiving any income in retirement outside of your portfolio, and in light of inflation-adjusted SPIAs going away, would you still recommend VPW?

Or perhaps a very conservative SWR instead?

Silent cal, I don't understand the context of your question.

According to an earlier post, your couple will be getting $45,000/year (after haircut) from Social Security. Why would you ignore this in your planning?

The wiki page says:

VPW is best used in conjunction with guaranteed base income from Social Security, a pension (if any), and (if necessary) a SPIA (possibly with annual cost-of-living adjustments).

The key word is "possibly". There's no requirement for cost-of-living adjustments. The VPW worksheet takes into account pensions and SPIAs with and without cost-of-living adjustments. Maybe you had read an earlier version of the wiki page. The change was discussed in October 2023.

Sorry if this was a confusing question. I am personally factoring Social Security into my VPW spreadsheet. I was just curious if "VPW is best used in conjunction with guaranteed base income from Social Security, a pension (if any), and (if necessary) a SPIA (possibly with annual cost-of-living adjustments)" implied VPW is only an appropriate tool if you have guaranteed income, or if you'd still be comfortable using it without. This is primarily in the context of younger people in my age cohort pessimistic about the future of Social Security or pensions and preferring to plan with the assumption they'll get nothing, and being pleasantly surprised if they do receive benefits in the future.

I imagine if you could live off off the 'Required Flexibility' amount VPW would still be an okay tool to use? Or perhaps taking the approach you've mentioned in earlier threads of taking the new portfolio balance in the 'Required Flexibility' tab and inputting it as the new Portfolio Balance to essentially model the effect of a 50% market drop followed by another 50% market drop.

[azanon]

Very helpful thank you.

I have another unrelated question. If you’re not planning on receiving any income in retirement outside of your portfolio, and in light of inflation-adjusted SPIAs going away, would you still recommend VPW?

Or perhaps a very conservative SWR instead?

Silent cal, I don't understand the context of your question.

According to an earlier post, your couple will be getting $45,000/year (after haircut) from Social Security. Why would you ignore this in your planning?

The wiki page says:

VPW is best used in conjunction with guaranteed base income from Social Security, a pension (if any), and (if necessary) a SPIA (possibly with annual cost-of-living adjustments).

The key word is "possibly". There's no requirement for cost-of-living adjustments. The VPW worksheet takes into account pensions and SPIAs with and without cost-of-living adjustments. Maybe you had read an earlier version of the wiki page. The change was discussed in October 2023.

Sorry if this was a confusing question. I am personally factoring Social Security into my VPW spreadsheet. I was just curious if "VPW is best used in conjunction with guaranteed base income from Social Security, a pension (if any), and (if necessary) a SPIA (possibly with annual cost-of-living adjustments)" implied VPW is only an appropriate tool if you have guaranteed income, or if you'd still be comfortable using it without. This is primarily in the context of younger people in my age cohort pessimistic about the future of Social Security or pensions and preferring to plan with the assumption they'll get nothing, and being pleasantly surprised if they do receive benefits in the future.

I imagine if you could live off off the 'Required Flexibility' amount VPW would still be an okay tool to use? Or perhaps taking the approach you've mentioned in earlier threads of taking the new portfolio balance in the 'Required Flexibility' tab and inputting it as the new Portfolio Balance to essentially model the effect of a 50% market drop followed by another 50% market drop.

It's hard to imagine any scenario where fans of VPW would not recommend VPW for retirement income. Just about everyone can create that scenario where it's "best used" since almost everyone will have social security, and if you don't have a pension, then you can "buy" a SPIA. If you're worried about volatility in monthly, quarterly, or annual payments being too high because of no pension, and maybe even no SPIA too, there are other levers to pull, such as using a lower stock allocation. For example, going from a 60/40 to a 30/70 all by itself should compensate for the lack of a pension vs. someone else that had one. And don't feel bad about doing that since a 30/70 is much closer to a minimum variance portfolio anyway. If I were comparing things to be afraid about, I'd be far more inclined to be afraid using a stock dominated portfolio after I retired than social security completely going away.

Hard to imagine a scenario to not recommend VPW because the "withdrawal efficiency ratio" of VPW is going to be outstanding regardless, because it spends into principle to fund retirement. Any method that doesn't do that will be less efficient (that's most of them such as "4% rule".)

[CalPoppy]

When an amount is entered into the "Temporary Retirement Income" section(s) on the "Retirement" sheet of the VPW Accumulation And Retirement Worksheet, what does the worksheet assume the retiree does with that amount? Invest it all into their portfolio?

[alterEcho]

When an amount is entered into the "Temporary Retirement Income" section(s) on the "Retirement" sheet of the VPW Accumulation And Retirement Worksheet, what does the worksheet assume the retiree does with that amount? Invest it all into their portfolio?

I believe the assumption is that you would be spending that money instead of drawing down for that part of your retirement income. I would have to look at the calculations, but it seems to reduce the value in the "Portfolio Withdrawal" calculated field and several other fields change based on the recognition of future income.

[Raspberry-503]

I know the 50% stock market devaluation for the required flexibility is somewhat arbitrary and the calculations assume that bonds stay put (wasn't the case in 2022) but I was trying to understand if it's reasonable, i.e. is it a once in a decade or once in a lifetime event?
I'm using some historical market data that either comes from the old VPW back testing spreadsheet from Siamond's sheet

That data shows:

Code: Select all

YEAR	 MARKET RETURN
1928		1.466628637
1929		0.91230893
1930		0.781206817
1931		0.570713392
1932		0.914457831
1933		1.548660085

so stock at $1000 at the start of 2029 was worth 1 x 0.91230893 x 0.781206817 x 0.570713392 x 0.914457831 = $372 by the of of 1932. that's about a 63 % drawdown. I think a monthly chart would show an even deeper dip but I don't have that data handy.

Mining the full historical data table from 1871 to 2023 I get this graph:


Interestingly the vertical scale is quite a bit off from the graph in https://awealthofcommonsense.com/2018/0 ... drawdowns/


Is my data or calculation incorrect? My data is "entire US" whatever that meant in 1926, whereas the data on that site is called SP500 drawdowns (but I don't think the SP500 existed in 1926?). My calculations use the method I described above for the 1930s crash and I think that's correct. Maybe it's because their graph shows monthly data?

But seeing numbers close to 90% drawdown seem to indicate a 50% crash is a once in a decade occurrence and a double-crash is more like once in a century.

The charts above are in nominal dollars.

In its simplest form, VPW just tells you in now-dollars how much you can take from today's portfolio value and things like inflation don't have a bearing on that, but I think inflation does matter. What really makes 1966 a bad year to retire is not so much bad market performance but rampant inflation for a decade. At least inflation was negative in the early 1930s

Does it make sense to look at inflation-adjusted drawdowns? Is that a thing? I can compute and graph it:


Wow that makes things interesting! fatter and deeper troughs than in the top chart. Suddenly a drawdown to -50% or at least -40% in "real dollars" seems like a more regular occurrence that is likely to last longer. Having the Required Flexibility to accommodate a 50% equity drop definitely seems wise.

Is my thinking flawed? Either way, the 50% drop of equity value flexibility seems like a pretty good rule of thumb, but if the last graph is meaningful, then doing a double-whammy of feeding the 50% back into VPW to do a "stress test" seems too extreme (never happened in past history), but of course if your portfolio (really your flexible lifestyle) can take it, the more power to you.

[Raspberry-503]

Another graph based on my historical data table. It tracks the drawdowns for for a portfolio of 70% stocks, 30% of which are International, and 30% bonds. I takes care of showing the effect of both 2022 when both stocks and bonds went down.
There are times where returns are not so dampened by bonds, so the Required Flexibility may need to be a little higher (lower income) than calculated by the tool for bond-heavy AAs.

This is the "inflation adjusted" flavor of the chart, similar to the bottom chart in my previous post.


Maybe I shouldn't have included international returns in this chart, not sure how much impact they have, the tables I use don't have international data before 1972, so they use US data instead for those years.

[CalPoppy]

When an amount is entered into the "Temporary Retirement Income" section(s) on the "Retirement" sheet of the VPW Accumulation And Retirement Worksheet, what does the worksheet assume the retiree does with that amount? Invest it all into their portfolio?

I believe the assumption is that you would be spending that money instead of drawing down for that part of your retirement income. I would have to look at the calculations, but it seems to reduce the value in the "Portfolio Withdrawal" calculated field and several other fields change based on the recognition of future income.

Thank you

[Harmanic]

I'm several years away from retirement and using VPW... But my working model was to use VPW as my "total net income" - much like my current paycheck. If we don't spend the entire amount, it just goes into our savings. So to my example above, that "unspent" $3k is now just part of the portfolio - it's no longer available to withdraw separately* - but it impacts the VPW result going forward. (* In truth, one of the appeals of VPW to me is it's dynamic and self-correcting nature. If we needed/wanted to spend a one-time extra $5k amount - I wouldn't be stressed by doing so... Much like my paycheck today - if we need to spend more on a given month, we know we'll need to make that up on other months...)

Could you just earmark this for a contingency fund for major expenses and keep carrying the excess over to account for things like a new car, roof, etc.? Behaviorally, if one just pulled funds from the portfolio as needed, it seems like it could become a habit and lead to a quicker depletion of funds.

[stuper1]

I'm several years away from retirement and using VPW... But my working model was to use VPW as my "total net income" - much like my current paycheck. If we don't spend the entire amount, it just goes into our savings. So to my example above, that "unspent" $3k is now just part of the portfolio - it's no longer available to withdraw separately* - but it impacts the VPW result going forward. (* In truth, one of the appeals of VPW to me is it's dynamic and self-correcting nature. If we needed/wanted to spend a one-time extra $5k amount - I wouldn't be stressed by doing so... Much like my paycheck today - if we need to spend more on a given month, we know we'll need to make that up on other months...)

Could you just earmark this for a contingency fund for major expenses and keep carrying the excess over to account for things like a new car, roof, etc.? Behaviorally, if one just pulled funds from the portfolio as needed, it seems like it could become a habit and lead to a quicker depletion of funds.

I'm still a few years from retirement myself, but my current plan is to take the total VPW suggested income amount out each month or quarter and put it into a separate money market fund that will earn interest but be safe from stock drawdowns. This will be the fund where I will spend from for my monthly expenses and pre-save for major upcoming expenses like a new car, home renovations, or a big trip. If the amount in the fund ever gets too large (say over $50,000 or so), then I might put some of that money back into my stock/bond investments.

[4nursebee]

Kudos to Longinvest for curating this thread.

I just got notice of a new reply to a post I made on this thread. It was really LI fixing broken links in something four years ago…